Institutional-grade scoring for retail alternative investment platforms.
Every platform is scored across four orthogonal pillars — Operational Risk, Investment Merit, Transparency & Alignment, and Liquidity & Access — combined into a 0–100 institutional composite with a letter grade, peer-group percentile, confidence indicator, and outlook flag. Every scored field is anchored to primary SEC filings with verbatim quotes and page references.
Four pillars
Separate scores for operational risk, investment merit, transparency, and liquidity
0–100 composite
Institutional scale calibrated for real dispersion across platforms
Peer percentile
Rank within asset-class peer group on each pillar and composite
Confidence & outlook
Honest indicators of data completeness and trajectory
Why four pillars and not one number?
A single composite collapses meaningful tradeoffs. A platform can have strong verified returns despite governance concerns; a different platform can have pristine governance with mediocre outcomes. Conflating these into one number tells you nothing actionable. Separate pillars preserve the nuance an investor or advisor actually needs to evaluate fit for a specific allocation.
Evidence Stack
Every score is anchored to primary sources.
AltStreet scores are produced from structured per-platform dossiers combining SEC EDGAR filings, offering documents, audited financials, and public regulatory records. AI-assisted extraction structures the data; human analysts verify material claims with verbatim quotes and page references. Every scored field is traceable back to the document it came from — and labeled with the analyst's confidence in the extraction.
Primary Source Documents
SEC EDGAR filings (Form D, 1-K, 10-K, amendments), offering circulars, audited financial statements, Form CRS, broker-dealer disclosures, and fund organizational documents — collected before scoring begins.
AI-Assisted Extraction
AI structures filings into normalized fields, surfaces discrepancies between marketing materials and SEC documents, and flags potentially material disclosures for analyst review. The analyst makes the final call on every scored field.
Human Verification
Material claims are verified against the underlying document. Each scored field stores a verbatim quote, page reference, source section, and an analyst-verified flag. The verification rate directly feeds the confidence indicator.
The evidence layer is the moat.
Most platform comparison sites cite marketing pages or platform-issued press releases. AltStreet scores are built from the documents platforms file with the SEC — including the disclosures their marketing pages omit. Every score is defensible because the source document is one click away.
Four-Pillar Architecture
Four orthogonal pillars, each scored independently.
Every platform receives four pillar scores plus a composite. Each pillar measures a distinct dimension of platform fitness; strength in one cannot mask weakness in another. This is the architectural decision that separates AltStreet from single-number scoring sites — and the structural difference between a forensic analysis and a marketing summary.
Operational Risk
Measures platform structural integrity. Governance, audit independence, administrator financial health, regulatory compliance history. A low score here means the platform has structural concerns that warrant scrutiny regardless of how the underlying investments perform.
Inputs
- · Governance structure and conflict-of-interest disclosure
- · Independent audit and administrator presence
- · Administrator cash position and going-concern signals
- · Prior regulatory enforcement and compliance history
- · Reporting quality and material event disclosure
Audit Vehicle Principle
Audit and administrator quality are classified based on the investment vehicle the investor actually holds, not only the parent operator or a separate pooled fund. AcreTrader is the model case: Proterra AcreTrader Farmland Fund LP is audited by Ernst & Young LLP and administered by CliftonLarsonAllen LLP, but individual farm-level LLCs used for Reg D single-property offerings do not disclose independent audits at the LLC level. Where assurance exists at the pooled fund level but not the specific offering LLC, AltStreet records the relationship as tiered and notes the limitation in audit/admin quality notes.
When a platform operates across multiple product cohorts - current vs. historical, originated vs. distributed - and audit or administrator relationships vary by cohort in ways that cannot be uniformly classified at the platform level, the appropriate classification is "unknown" with detailed notes. That is honest rather than evasive: forcing a tier classification when the evidence does not support one would convey false precision.
Investment Merit
Measures actual investor outcome quality. Realized IRR across verified exits, exit success rate, target-vs-actual tracking error, sample maturity. Strong realized returns lift this pillar even when operational concerns drag others — this is the pillar that captures whether the platform has actually delivered for investors.
Inputs
- · Median realized IRR across verified exits
- · Exit count and sample maturity (curated subset detection)
- · Loss rate on exited offerings
- · Target-vs-actual IRR tracking error
- · Asset-class-aware return benchmarking
Transparency & Alignment
Measures the investor's information position. How estimable are total economics from disclosed documents? Are sourcing fees, performance fees, and carry verifiable from balance-sheet primitives? Are related-party flows disclosed? This is the pillar that captures whether an investor can actually model what they're buying.
Inputs
- · Fee transparency — estimability from disclosed primitives
- · Affiliate fee complexity and related-party flows
- · Disclosure completeness (PPM depth, risk-factor specificity)
- · Tax document type and delivery timing
- · K-1 vs 1099 classification and complexity
Liquidity & Access
Measures real-world investor access and exit optionality. Secondary market existence, active trading volume vs. theoretical optionality, redemption fulfillment rates, minimum investment thresholds. This pillar separates platforms with functional secondary markets from those with theoretical optionality.
Inputs
- · Secondary market existence and active trading volume
- · ATS registration and broker-dealer infrastructure
- · Redemption fulfillment rates and gate provisions
- · Minimum investment thresholds and access tiers
- · Realistic time-to-exit based on historical patterns
Why this matters in practice. A fractional art platform might score B on Investment Merit (strong verified returns) while scoring C- on Operational Risk (administrator stress) and D on Liquidity & Access (thin secondary market). A single composite would obscure that profile; four pillars make it legible. An RIA can see the platform has produced for investors but warrants attention to operational and liquidity factors. That's the actionable read.
Score Anatomy
What's in a score.
Beyond the four pillars, every platform receives five additional outputs that make the composite legible. Together they allow a user to assess not just where a platform stands, but how confident that assessment is, where it ranks against direct peers, and which way it is trending.
Composite
0–100
Weighted combination of the four pillars. Most platforms cluster 50–75; the top decile (80+) and the tail (under 40) are deliberately rare.
Letter Grade
A+ – F
Standard ratings-system overlay for at-a-glance comparison. Composite and each pillar receive independent letter grades.
Peer Percentile
0–100
Rank within the platform's defined peer group at both the composite level and within each pillar. The relative number an RIA actually uses for allocation decisions.
Confidence
5 levels
Very high to very low. Driven by data completeness, evidence verification rate, source quality, and staleness.
Outlook
↗ → ↘ ⚠
Improving, stable, declining, or deteriorating — computed from trailing score history and material change events.
Letter grade calibration
Calibrated for institutional dispersion — most platforms cluster in B/C+ territory; A and A+ are reserved for platforms with excellence across all four pillars and no material flaws.
A+
90+
A
85
A-
80
B+
75
B
70
B-
65
C+
60
C
55
C-
50
D+
45
D
40
D-
35
F
< 35
Audience-aware composite weighting
The four pillars combine into the composite using a weighting appropriate to the audience. Balanced for general use, investor weighting (Investment Merit + Liquidity emphasized) for return-focused use, operational weighting (Operational Risk + Transparency emphasized) for compliance-focused RIA use, institutional weighting for due-diligence use. Same pillar scores, different composite for different decision contexts.
Peer percentile per pillar
Beyond the composite peer percentile, each pillar has its own peer-relative rank. A platform might rank 1st on Investment Merit within fractional art but 4th on Operational Risk — meaningful information for an RIA evaluating which platform to use for which client allocation.
Aggregation Methodology
Different aggregation methods, applied where each fits.
Most scoring systems use a single weighted average across all components. That approach collapses dispersion — one bad value gets averaged out by strength elsewhere, and platforms converge to the middle. AltStreet uses three different aggregation methods within and across pillars, depending on whether a component is a bottleneck, substitutable, or additive.
Harmonic mean
Used for dimensions where one bad value should disproportionately tank the score because investors literally cannot underwrite around it. Fee transparency, reporting quality, and audit independence are bottlenecks — no amount of strong track record substitutes for the inability to model fees.
Geometric mean
Used for dimensions where strength in one area partially compensates for weakness in another, but no single dimension can be near zero. Track record and realized returns are partially substitutable within the Investment Merit pillar — a strong verified IRR can offset modest qualitative track record signals.
Arithmetic mean
Used for dimensions that are genuinely additive — qualitative liquidity assessment combined with quantitative trading reality, where strength in one is straightforwardly substitutable for strength in another and no specific dimension is a dealbreaker on its own.
Excellence-rare scaling
The composite is scaled to the 0–100 institutional range using a deliberately non-linear curve calibrated so that achieving 90+ is meaningfully harder than achieving 50+. A composite of 70 represents strong performance across all four pillars. A composite of 85 requires excellence across most pillars and no material flaws. A composite of 90+ is reserved for platforms that excel across every pillar and have human-verified primary sources for every material claim. This preserves dispersion at the top end — the level where allocation decisions actually depend on small score differences.
Asset-class–aware return thresholds
Investment Merit is scored against exposure-type-specific benchmarks. A 12% loan interest target is not equivalent to a 12% equity-appreciation target. Loan interest, equity appreciation, rental income, revenue share, and NAV appreciation each have their own threshold curves. This is what allows cross-asset-class comparison to work: a 12% loan and a 12% equity IRR produce different Investment Merit scores because they represent different risk-adjusted outcomes.
Sample maturity adjustment
Realized return data is adjusted for sample maturity. A platform that has exited 5% of its portfolio with a strong median IRR gets less credit than a platform that has exited 50% with a comparable median, because the curated subset effect is real: early exits tend to be the strongest performers. The adjustment is proportional — full credit at 50%+ maturity, modest discount at 25–50%, larger discount under 25%, significant discount under 10%.
Floor Caps
Categorical disqualifiers cap relevant pillars.
Some platform attributes are categorical disqualifiers — issues so severe that no amount of strong performance elsewhere should produce a high score in the affected pillar. Floor caps enforce this. When a floor-cap condition is detected, the relevant pillar is capped at a specific tier ceiling. Caps are surfaced explicitly alongside the score so users can see exactly which issue triggered them.
Severe reputational risk
Documented severe reputational issues with platform principals (active SEC enforcement actions, criminal charges, material civil judgments) cap the relevant pillar at D tier.
Affects: Operational Risk
Active regulatory enforcement
Open investigations or enforcement actions against the platform itself cap the Operational Risk pillar at D+ tier.
Affects: Operational Risk
Opaque affiliate fee structures
Multi-layer fee arrangements where related-party flows cannot be reconstructed from filed documents cap the Transparency & Alignment pillar at C tier.
Affects: Transparency
Severe deal-level reputational risk
Deal-level: severe reputational risk on deal principals caps the deal score at D tier regardless of underlying economics.
Affects: Deal Score
Pre-construction with no collateral
Deal-level: pre-construction offerings without collateral existing at close cap the deal score at C+ tier.
Affects: Deal Score
Multi-layer opaque fees (deal)
Deal-level: multi-layer opaque fee structures cap the deal score at C tier.
Affects: Deal Score
Floor caps are reserved for true disqualifiers.
Resolved compliance issues, going-concern qualifications, moderate reputational concerns, and similar material-but-not-fatal disclosures are penalties, not caps. They reduce the pillar score but don't impose a ceiling. This is intentional: a platform with a going-concern flag on its administrator but a strong realized return profile shouldn't be capped in the Investment Merit pillar — that's exactly what pillar separation is for. Caps are reserved for cases where the issue is structurally fatal to the dimension being measured.
Floor caps are surfaced, not hidden. When a floor cap is applied, the score-before-cap value is preserved and the applied cap reason is shown explicitly. Users can see exactly why a pillar was capped and read the underlying evidence that triggered the cap. This makes the methodology auditable: if a cap feels wrong, the user can examine the source document themselves.
Peer Groups
Scores are most useful relative to direct peers.
Comparing a fractional art platform to a farmland platform on the same absolute scale is technically possible but rarely useful. An RIA evaluating where to allocate a client's art-sleeve capital cares about how a platform ranks against other art platforms — not against farmland. AltStreet produces peer-relative scoring as a first-class output: every platform's composite and pillar scores are published alongside their ranks and percentiles within their defined peer group.
Per-pillar peer rankings
Each pillar has its own peer-relative rank within the peer group. A platform can be #1 in Investment Merit within its peer group but #4 in Operational Risk — meaningful information for an RIA matching platforms to specific client risk tolerances.
Peer groups evolve as coverage grows.
Some peer groups today have 3–5 platforms; others have 10+. As coverage expands, peer groups become statistically more meaningful and percentile rankings carry more information. Peer-group taxonomy is reviewed quarterly and refined as the platform universe evolves.
Confidence & Outlook
A score is only as useful as the data behind it.
Every score is published with two supporting indicators that protect against false precision: a confidence level reflecting how complete and verified the underlying data is, and an outlook flag reflecting score trajectory. Both are first-class outputs — visible alongside the score wherever the score is displayed.
Confidence indicator
Reflects how confident AltStreet is in the score itself. Computed from data completeness, evidence verification rate, source quality, and staleness.
Outlook trajectory
Reflects score trajectory based on trailing score history and material change events. A platform's score isn't just where it is; it's where it's heading.
Driver Attribution
Every score shows what's driving it.
The composite score is the headline. The drivers are the explanation. For every platform, the top positive and top negative contributors are surfaced explicitly — with the dimension name, the magnitude of contribution, and the specific evidence detail that produced it. An RIA can see not just that Masterworks scored 61, but that the score reflects strong verified returns and fee transparency offset by administrator financial stress and thin secondary market liquidity.
Top positive drivers
The three dimensions contributing most to the platform's composite. Surfaces which pillars and which inputs within pillars are lifting the score. Each driver shows its magnitude in score points.
Top negative drivers
The three dimensions pulling the composite down most. Surfaces which penalties, which pillar weaknesses, and any applied floor caps. Each driver shows its magnitude and the underlying evidence detail.
Why this matters for institutional users. When an RIA recommends a platform to a client, they need to be able to explain why. "AltStreet scored this 72/B+ with strong Investment Merit and Operational Risk offset by moderate liquidity concerns" is a defensible statement; "AltStreet rated it highly" is not. Driver attribution turns the score from a black-box number into an auditable analytical product.
What AltStreet scoring is — and isn't.
Scoring IS
- A four-dimensional comparison framework for evaluating platform operational risk, investment merit, transparency, and liquidity independently.
- Anchored to primary sources — every scored field traceable to a specific document with a verbatim quote and page reference.
- Designed for institutional dispersion — the math produces meaningful differentiation across the platform universe rather than collapsing to the middle.
- Continuously updated — scores refresh as new filings arrive; material changes trigger score change events and outlook adjustments.
Scoring IS NOT
- A buy or sell signal. Scores measure structural and merit dimensions, not market opportunity or suitability for any specific investor.
- A prediction of future returns. Realized track record is one input; future performance is not the output.
- A substitute for personal due diligence. Scores summarize evidence; investors and advisors must still evaluate fit against their specific constraints.
- Influenced by relationships. Platforms cannot pay for higher scores or removal of material findings. The methodology is identical for every platform covered.
Coverage by asset class
The four-pillar methodology applies across asset classes. Category hubs carry the full directory of platforms covered; the index below links to each category's hub page.
Known limitations
Private-market research has real evidence limits. AltStreet names those limits rather than smoothing them into false certainty.
Private markets are opaque
Many platforms disclose structure, fees, and eligibility more clearly than realized outcomes, asset-level marks, or failed offerings.
Performance claims vary in verifiability
Reported IRRs, target yields, and track records may depend on platform-selected samples unless exits can be tied to primary documents.
Exit data is incomplete
Not all platforms publish realized exits, defaults, hold periods, buyer types, or distributions at the offering level — limiting Investment Merit scoring.
Sample maturity matters
Even platforms with disclosed exit data may have low sample maturity — early exits tend to be the strongest performers, creating a curated subset effect that the methodology adjusts for.
Peer groups are imperfect
Some platforms span multiple peer groups (e.g., a multi-asset-class platform). Peer assignment reflects primary asset focus; secondary classifications may be added over time.
Regulatory facts can change
Broker-dealer status, offering exemptions, custody relationships, and platform disclosures are checked against dated source material. Scores update when material changes occur.
Independence and disclosure
Methodology independence: Scoring is produced using the same methodology for every platform. Platforms cannot pay for higher scores, removal of material findings, or preferential placement. The same penalty values, threshold tables, and floor-cap rules apply regardless of any commercial relationship.
Affiliate compensation: AltStreet may receive affiliate compensation when readers click links to platforms and subsequently invest. Affiliate status does not change the scoring methodology, source standards, or analytical findings.
Methodology versioning: The scoring methodology is versioned. When the methodology is updated, the version is recorded alongside every score, and historical scores remain accessible for reference. This allows users to see whether a score change reflects a platform change, a methodology change, or both.
Right to respond: Platforms are welcome to request review of material findings. Where a platform provides additional primary-source evidence that materially changes a scored field, the score is updated and the change is logged. Scores are not changed in response to general disagreement absent new primary-source evidence.
See the methodology applied.
Browse the comparison terminal to see four-pillar scoring applied to live platform data, or read full platform reviews that walk through the underlying evidence behind each score.
Platform evaluations are educational resources and do not constitute investment advice or recommendations. Always conduct independent due diligence and consult qualified professionals before making investment decisions.